INDEX

Explanations

phrases related to a negative attitude towards certain behaviors or individuals

oai_token-act-pair · gpt-3.5-turbo

phrases that express respect or concern toward various subjects

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

breaks

-0.72

quad

-0.71

gallery

-0.71

NAS

-0.69

edin

-0.69

ns

-0.69

wcs

-0.69

Tracker

-0.68

sts

-0.67

daq

-0.66

POSITIVE LOGITS

 conformity

0.86

 oneself

0.77

 mortality

0.74

 rationality

0.71

 autonomy

0.70

 criminality

0.68

quo

0.65

 populated

0.65

 labour

0.65

 antiqu

0.64

Activations Density 0.256%

phrases related to a negative attitude towards certain behaviors or individuals

phrases that express respect or concern toward various subjects

No Comments

No Known Activations